I couldn't wait for the MDBT42V Espruino, so following guidance from Gordon, flashed one myself and made a little PCB and ring mount for it - the ring is currently bigger than it should be because it'll eventually have micro-USB charging.
It's mostly so I have a way to press buttons when using hand tracking in flight sims, but it turns out it's good for other things too (see video).
Hey, a Worlds First . . . .
a directional Point-n-Shoot bit flinger! (see video #1 post)
thats class! I assume you have IMU and a button there. Whats the battery life?
No IMU - finger tracking is through an Ultraleap. I'm hoping 2-antenna UWB becomes a readily-available thing faster though.
Not sure about battery life yet, because I keep making mistakes faster than my testing intervals, but so far it looks like its 11mAH battery can give about a day of paired usage with moderate amounts of clicking (I also use it as my mute button in all my VC meetings).
Ok, if its just the button battery life shouldn't be an issue. Ultraleap seems a bit short range but probably perfect for your application. TOF cameras are more available and cheap nowadays. I havent tried it for gestures though.
That's amazing! Thanks for posting up!
It's always been the battery that seemed tricky in tiny devices, but actually those MS920SE/ML-920S seem really promising. Did you consider wireless charging? It might be that a few loops of copper around the ring and then a diode (and 3.1v zener to clamp the voltage) would be enough to charge the battery - and it might let you save some space.
I've been wondering if I should start stocking pre-programmed MDBT42Vs? They're really cool but you definitely need a hot air gun and paste to get them soldered!
Inductive is a cool idea - I'll take a look, thanks! I would love in the fullness of time to do magnetic/pogo pins/pads and something like a Grepow 160725 curved battery too.
Soldering the MDBT42V is tough - with my hotplate, hot air gun, and total-amateur skills, it took several attempts to get it connected properly - I eventually found that I needed to apply a bit of downwards pressure to it while soldering to get all the contacts to connect - I guess because my solder paste wasn't evenly distributed so there was a little lift (turns out Raytac even warn about this).
Not sure you want to sign up for all the bug reports from the consequences of similarly tough soldering situations, but I would definitely buy a few if you stocked them - especially since I bought all the remaining stock Amazon had :)
Yes - I think they'd have to come with a warning about 'solder at your own risk'!
I actually found the MDBT42V surprisingly easy to solder - I've done about 25 so far and every single one worked (which is far better than if I manually soldered the MDBT42Q!). However I did always press them downwards while the solder was melted, so maybe that's the trick to getting them working!
Despite my first success, over the past 24 hours I've failed at five of them - it could be my PCB design, but regardless of whether I use a hotplate or hot air, I invariably end up either not getting a connection or delidding the MDBT42V. Any tips? Are you using a solder stencil? Very very precisely pasting each pad? Smearing paste on both and heating?
Did you try to put solder on the MDBT42V before actually soldering it on the PCB?
I soldered 5x5 LFCSP footprints using generic metal stencil and cheap reflow oven. I would look for a generic stencil for that footprint or design one and fabricate it. OSH makes stencils, I made some with them oshstencils. I never tried plastic, they are super cheap. Reflow oven you can even build by yourself. Maybe fast option would be to find some fabrication house nearby.
With ICs and passives setting profiles is easy. It might be different story with modules. The heating profile might be different.
I actually just smeared mine all over with solder paste, then pushed it down with tweezers when hot so the excess spilled out the edge as little balls. Then when cool you just run around the edge and the balls drop off.
I was using leaded solder, but I guess if you're using normal temp unleaded paste you might need to get it so hot before the paste melts that the module has issues? You could try getting some low temperature unleaded paste to try - as far as I can tell there seems to be very little downside to using low temp paste.
I've heard of folks laser-cutting stencils out of mylar with standard laser cutters, so you could always give that a go? But again, I never actually bothered with that.
Thanks all - I do have a stencil, but the 2D array of pads makes it much tougher than the edge-contact ones I've tried before (can't visually inspect, and can't fix with a soldering iron), so a lot of my issue is "how do I know if I've got it right".
Gordon's method ended up being the most successful - the squeezed-out solder balls inspired a lot of terror ("if solder is coming out the side, might it not be squishing across pads internally?"), but everything seems to be working. To mitigate my own terror, in my next version of the PCB I've also omitted the pads I'm not using, which is almost certainly a mistake for a variety of reasons, but its made layout much easier and I can only learn by doing :)
Attached is a shot of the latest version, with Micro-USB charging. I swapped some pins and setWatch isn't working on the ones I swapped it to (11, 28, 30) - might be a soldering issue again, but just in case the next version goes back to 0,2,3,4,5,29
what if the pad on the PCB is a via with the small hole, does the capillary force suck in the solder from the back side of the PCB?
Don't worry about formatting, just type in the text and we'll take care of making sense of it. We will auto-convert links, and if you put asterisks around words we will make them bold.
For a full reference visit the Markdown syntax.
© Espruino, powered by microcosm.
Report a problem